For Volume 1:

Geotechnical and Geophysical Site Characterisation 5 – Lehane, Acosta-Martínez & Kelly (Eds)
Geophysical properties
© 2016ofAustralian
soils Geomechanics Society, Sydney, Australia, ISBN 978-0-9946261-1-0
Geophysical For properties
J. C. Santamarina and J. Park
Volume 2: of soils
Earth Science & Engineering, King Abdullah University of Science and Technology (KAUST)
J. C. Santamarina Geotechnical
and J. Parkand Geophysical Site Characterisation 5 – Lehane, Acosta-Martínez & Kelly (Eds)
Earth Science & Engineering,©King
2016Abdullah
Australian Geomechanics
University Society,
of Science andSydney, Australia,
Technology ISBN 978-0-9946261-2-7
(KAUST)

ABSTRACT: Low energy perturbations used in geophysical methods provide insightful information about con-
stant-fabric soil properties and their spatial variability. There are causal links between soil type, index proper-
ties, elastic wave velocity, electromagnetic wave parameters and thermal properties. Soil type relates to the
ABSTRACT:
stress-dependent Low energyvelocity,
S-wave perturbations
thermalusedandin electrical
geophysical methods provide
conductivity insightful information
and permittivity. The small strain about stiff-
con-
stant-fabric soil properties and their spatial variability. There are causal links between
ness reflects the state of stress, the extent of diagenetic cementation and/or freezing. Pore fluid chemistry, fluid soil type, index proper-
ties,
phaseelastic wave velocity,
and changes in eitherelectromagnetic
fluid chemistrywave parameters
or phase manifest andthrough
thermalelectromagnetic
properties. Soilmeasurements.
type relates toThe the
stress-dependent S-wave velocity, thermal and electrical conductivity and permittivity.
volumetric water content measured with electromagnetic techniques is the best predictor of porosity if the water The small strain stiff-
ness reflects
saturation the state
is 100%. of stress,
Changes in the extent
water of diagenetic
saturation alter thecementation
P-wave velocity and/orwhen freezing. Pore fluid
Sr100%, chemistry,
the S-wave fluid
velocity
phase and changes in either fluid chemistry or phase manifest through electromagnetic
at intermediate saturations, and the thermal conductivity when the saturation is low Sr0%. Finally, tabulated measurements. The
volumetric water content measured with electromagnetic techniques is the best
values suffice to estimate heat capacity and latent heat for engineering design, however thermal conductivity predictor of porosity if the water
saturationmeasurements
requires is 100%. Changes underinproper
waterfield
saturation alter the P-wave velocity when Sr100%, the S-wave velocity
conditions.
at intermediate saturations, and the thermal conductivity when the saturation is low Sr0%. Finally, tabulated
values suffice to estimate heat capacity and latent heat for control the mechanical
engineering design, howeverbehaviour of a conductivity
thermal sand-gravel
requires measurements under proper field conditions.
1 INTRODUCTION mixture when the gravel is looser than e G
max
. Similar
analyses define the 13 notable mixtures listed in Ta-
ble control
1. The thegravimetric-volume
mechanical behaviour of a presented
equations sand-gravel in
Geophysical methods have played a crucial role in mixture when the gravel is looser than e G
max
. Similar
1subsurface
INTRODUCTIONcharacterization, in the detection of hydro- this table compute the corresponding fractions for
analyses
gravel, sand define
and the
fines.13Specific
notable factors
mixtures listed in
included in Ta-
the
carbon and mineral resources, and in monitoring a ble 1. The gravimetric-volume equations presented in
Geophysical methods have played a crucial
wide range of subsurface processes. Geophysical sur- role in definition of these mixtures reflect an extensive data-
subsurface characterization, this table compute the corresponding fractions for
veys are minimally invasiveinandtheinvolve
detection of hydro-
low-energy base of soil properties (details in Park & Santamarina
gravel, sand and fines. mixtures
Specific factors included in the
carbon and mineral resources, and in monitoring
stimuli, thus, they are inherently non-destructive con- a 2016). These notable determine classifica-
wide range of subsurface processes. Geophysical definition of these mixtures reflect an extensive data-
stant-fabric measurements. In addition, the samesur-ge- tion boundaries in the triangular RSCS classification
base of soil properties (details in Park & Santamarina
veys are minimally
ophysical assessmentinvasive and in
takes place involve low-energy
laboratory studies chart. Analyses and experimental data demonstrate
stimuli, thus, they are inherently non-destructive con- 2016).
that mixture Thesethresholds
notable mixtures
are differentdetermine
for flow classifica-
and me-
as in field surveys (Note: laboratory measurements tion boundaries in the triangular RSCS classification
stant-fabric measurements. In addition, the
may require careful analysis prior to the interpretation same ge- chanical control. Table 1 presents the classification
ophysical assessment takes place in laboratory chart. Analyses and experimental data demonstrate
of field data due to frequency-dependent wave studies
disper- procedure:
that mixture thresholds are different for flow and me-
as
sioninand
field surveys (Note: laboratory measurements
scattering).
may require careful analysis  Input
chanical the coefficient
control. Tableof1uniformity
presents the Cu and the mean
classification
Geophysical parameters areprior
mostto relevant
the interpretation
to engi- particle
procedure: roundness R, of gravel and sand fractions (or
of field data
neering due to
design. frequency-dependent
This manuscript presents wave disper-
a concise
sion and scattering). their values of e max
and e min
).
yet practical summary of the geophysical properties  Input
Input the
the coefficient
liquid limitofLLuniformity
of fines (or Cu the
andvoid
the mean
ratio
Geophysical parameters are most relevant
of soils. The four central themes include: soil classi- to engi- particle roundness R, of gravel and sand fractions (or
neering design. This manuscript presents a concise of fines at '=10max kPa, '=1minMPa, and at the liquid limit
fication, characterization with elastic waves, charac- their
LL). values of e and e ).
yet practical summary of the geophysical properties  Input the liquid LL of①-through-⑨
terization
of soils.
with
The
electromagnetic
four central themes
waves and
include: soil
thermal
classi- Compute notablelimit mixtures fines (or the void forratio
me-
properties. of fines at '=10 kPa,
chanical-control and '=1 MPa, and atforthe
⑩-through-⑬ liquid
flow limit
control;
fication, characterization with elastic waves, charac- LL).
terization with electromagnetic waves and thermal plot the 13 notable mixtures on the triangular chart
2 SOIL CLASSIFICATION andCompute
draw thenotable mixtures
classification ①-through-⑨
boundaries. Note: for the me-
Ex-
properties. chanical-control and ⑩-through-⑬ for flow control;
Step 1: Coarse or fine? Let’s accept the general defi- cel file available on the authors’ websites simultane-
plot
ouslythe 13 notable
draws the chart, mixtures on the triangular
classification boundarieschart and
nition
2 SOIL of CLASSIFICATION
gravel (retained on sieve #4), sand (passing and draw the classification boundaries. Note: the Ex-
through sieve #4 and retained on sieve #200) and plots the point that represents the soil under consider-
Step cel file available on the authors’ websites simultane-
fines1: Coarsethrough
(passing or fine?sieve
Let’s#200).
acceptAny the general defi-
one of these ation.
ously drawsthe thesoil
chart, classification boundaries and
nition of gravel (retained on sieve #4),
soil components can control the mechanical and hy- sand (passing  Classify under consideration. The double
through sieve #4 and plots the point that represents the soil under consider-
draulic behaviour of a retained on sieve
soil mixture. #200) anda
For example, letter nomenclature recognizes: first, the soil fraction
ation.
fines (passing through sieve #200). Any
densely packed sand near eS will carry the load and
min one of these that controls the mechanical behaviour and second,
soil components can control the mechanical and hy- theClassify the soil
soil fraction thatunder
controls consideration. The double
flow (in parenthesis).
draulic behaviour of a soil mixture. For example, a letter nomenclature recognizes: first, the soil fraction
densely packed sand near eS will carry the load and 135
min that controls the mechanical behaviour and second,
the soil fraction that controls flow (in parenthesis).
Table 1. Revised Soil Classification System RSCS.
Background:
Physics-based: gravimetric-volumetric analysis
Data-driven: takes into consideration extensive databases of soil behaviour

Input:
Gravel G (> 4.75mm) FG eGmax and eGmin or roundness R and uniformity Cu
Sand S (0.075~4.75mm) FS eSmax and eSmin or roundness R and uniformity Cu
Fines F (< 0.075mm) FF eF|10kPa, eF|1MPa , and eF|LL or liquid limit LL

Compute Threshold Fractions

Gravel fraction FG=MG/MT Sand fraction FS=MS/MT Fines fraction FF=MF/MT
1 1 1
 e eS eG   1  eS e   1  eS 1  eF 1  eF 
FG  1  G  
 FS   1 S 
 FF     1
 1  eS 1  eF 1  eS   G e 1  eF   eG eS eS 
Support information & Correlations:
flow
Fines: Fluid flow eF  0.05LL log(LL  25)
10 kPa 1kPa 1MPa 1kPa
Fines: Load carrying e F  eF  C c  0.026LL  0.07 eF  eF  3C c  0.011LL  0.21
0.154 0.522 0.082 0.371
Gravel and sand: eCmax  0.032   , eCmin  0.012  
R Cu R Cu
  0 Notable Mixtures
100 No.
10 Gravel Sand Fines
90 ① eGmin
- eF|10kPa
20

30 80 ② eGmin eSmax -
eGmin
eSmax
eF|10kPa
(a) Load carrying

70 ③
40
60 ④ - eS min
eF|10kPa
50
⑤ 2.5eGmax eSmin -
60 50
⑥ 2.5eG max
eS min
eF|10kPa
7 F(F) 40
70 8 ⑦ 1.3eGmax - eF|1MPa
80 9 30
⑧ - 1.3eSmax eF|1MPa
GF(F) SF(F)
1 20 ⑨ 2.5eGmax 1.3eSmax eF|1MPa
90
GSF(F) 6 4
10 3 S(F)
100 G(F)
GS(F)
10 ⑩ eGmin - λeF|LL
13
11 12
(b) Flow

G(G) GS(S) S(S)

0 ⑪ eGmin eSmax λeF|LL
2 5
0 10 20 30 40 50 60 70 80 90 100 ⑫ 2.5eGmax eSmin λeF|LL
Sand [%] ⑬ - eSmin λeF|LL

Classify the soil. Report the two-name nomenclature: first letter/letters indicate the load-carrying fraction the second
letter indicates the flow-controlling fraction. If either letter is F  proceed to classify the fines - Table 2.
Sources: Park and Santamarina (2016).

Step 2: Fines classification. Soils that are fines-con- Measured liquid limits are corrected for specific grav-
trolled -either in their mechanical and/or flow re- ity and precipitated salts. Then two ratios LLker/LLbrine
sponse- require further analysis to determine that type and LLDW/LLbrine combine the corrected values as
of fines. The most salient characteristics of fines are demonstrated in Table 2. The electrical sensitivity SE
(1) their specific surface, assessed by the liquid limit, captures the distance between measured values and
and (2) their sensitivity to pore fluid chemistry. We the absolute “non-sensitive” soil response at
determine fluid sensitivity by running liquid limits LLker/LLbrine=1 and LLDW/LLbrine=1. The two-letter
with deionized water LLDW, brine LLbrine to collapse pair classification of a fine soil recognizes its plastic-
the double layer (2 M NaCl solution), and kerosene ity and its electrical sensitivity (Table 2).
LLker to explore the effect of van der Waals forces.

136
Table 2. Fines Classification based on Plasticity and Sensitivity to Pore-Fluid Chemistry.
Background:
Specific surface and sensitivity to pore fluid chemistry are the salient characteristics of fines.

Input: Three liquid limits (soil fraction passes sieve #200 - fall cone method BSI 1990)
Sediment mixed with deionized water LLDW, kerosene LLker, and 2-M NaCl brine LLbrine.

Compute corrected liquid limit ratios

LL
1  cbrine brine
LLker LLker 100 LL DW LL DW  LLbrine 
 and  1  cbrine 
LLbrine corr LLbrine Gker LLbrine corr
LLbrine  100 
where Gker is the specific gravity of kerosene; cbrine=concentration of NaCl brine [mol/L]

Calculate the electrical sensitivity SE (use ratios above, or their inverse such that they are 1)

2 2 2 2
left  LLker   LLDW  ( right )  LL   LL DW 
SE    1    1 or S E   brine  1    1
 LLbrine   LLbrine   LLker   LLbrine 

3
Electrical Sensitivity SE
High SE

2 NH LH IH HH
Classify the soil

Report the two-letter pair for

 Plasticity 1
 Electrical sensitivity
Low Inter

NI LI II HI

NL LL IL HL
0
0 50 100 150 200
Non Low Inter High plasticity

Liquid limit using brine LLbrine (corrected)

Sources: Jang and Santamarina (2016a); Jang and Santamarina (2016b)

strain); reflection, refraction and scatter (heterogene-

3 ELASTIC WAVES ous); S-wave splitting (anisotropy); slow & fast
waves (mixed phase); Rayleigh, Love, and tube
A small-strain mechanical perturbation propagates waves (bounded media).
through the soil mass as an elastic wave. The wave
Wave velocity The shear stiffness of the granular skel-
equation for mechanical wave propagation combines
eton Gsk defines the S-wave velocity VS. Therefore, VS
Newton’s law F=m·a=m·d2u/dt2, dynamic force equi-
increases with effective stress, diagenesis (e.g., ce-
librium, constitutive equations (Hooke’s law =E),
mentation, creep, and salt precipitation), and suction
and compatibility of deformations. The resulting
(Table 3).
wave equation anticipates two modes of propagation
The P-wave velocity VP depends on the constrained
in a linear-elastic, homogeneous, isotropic, single-
modulus of the soil Msoil, which is a function of the
phase and infinite continuum: (1) S-waves where the
shear stiffness of the skeleton Gsk (see above) and the
particle motion is normal to the direction of propaga-
bulk stiffness of the soil Bsoil. The bulk stiffness of the
tion, and (2) P-waves where the particle motion is in
soil can be computed by successive substitutions to
the direction of wave propagation. New forms of
capture (Table 3): the skeleton stiffness Bsk and po-
wave propagation emerge anytime the above assump-
rosity n, the stiffness of water and air Bw and Ba, and
tions are relaxed, for example: shock waves (large
the degree of saturation Sr as presented in Table 3.

137
VP/Vs - Poisson’s ratio. Theory of elasticity antici-
pates that the ratio VP/VS is a function of Poisson’s
ratio  (Table 3). For small-strain wave propagation,
Poisson’s ratio is =0.150.05 for dry or unsaturated
soils, but it tends to 0.5 for well saturated soils;
for example, Poisson’s ratio can be =0.485 for a sat-
urated sand 20m deep and exceed =0.499 for a satu-
rated soft clay.
Attenuation. The wave amplitude decreases with dis-
tance due to (Table 4): geometric spreading of the
propagating spherical or cylindrical front (i.e., same
energy across a larger area), reflection and backscat-
ter at interfaces (lower transmitted energy), and unre-
covered energy consumed while deforming the mate-
rial as the wave propagates (viscous, thermoelastic
and/or frictional losses). Geometric spreading van-
ishes in plane waves.

Table 3. Elastic Waves: Velocity.

S-wave VS

saturated or dry soils; G sk  '    
VS     mean  where   0.73  0.27 log 
uncemented   1kPa  m/s
 
'mean mean effective stress on the polarization plane

 suction  S r 
unsaturated soils VS  VS  for Sr 1.0  1  
 0.75  ' v 
M soil Bsoil  4 / 3G sk
P-wave VP VP   water VP=1482 m/s
 soil  soil air VP= 343 m/s

Bulk modulus and mass density (Assumes Bsk/Bg  0 and low frequency limit)
2(1   )
Skeleton Bsk  G sk Gsk from Gsk  VS2 
3(1  2 )

Soil (fluid + skeleton) Bsoil  Bsus  Bsk  soil   sus  1  n  g  n fl

1
Suspension 1 n n 
Bsus    sus  1  n  g  n fl
(fluid + particles)  Bg B fl 


S 1  Sr 
1  fl  1  S r  a  S r  w
Fluid Mixture B fl   r  

B
 w Ba   fl  S r  w in case aair

VP M soil 21    unsaturated soils =0.150.05

VP/VS – Poisson’s ratio  
VP G sk 1  2  well saturated soils 0.5

Sources: Richart et al. (1970); Hardin and Drnevich (1972); Reynolds (1997); Santamarina et al. (2001); Cho and Santamarina (2001).
Based on an earlier compendium by Santamarina et al. 2005.

rameters combine to determine the propagation ve-

4 ELECTROMAGNETIC WAVES
An electrical or magnetic transient propagates
through the soil mass as an electromagnetic wave.
The wave equation for electromagnetic waves results
from Maxwell’s equations. The electrical field and
the magnetic field oscillate transversely to the direc-
tion of propagation and to each other.
There are three electromagnetic material parame-
ters: magnetic permeability μ, electrical conductivity
σel, permittivity κ. Permittivity and permeability are
complex quantities because they represent both in-
and-out of phase responses. The three material pa-
locity V and the material attenuation coefficient α or
skin depth Sd =1/α i.e., is the distance travelled by a
plane wave when its amplitude decreases to 1/e of the
initial amplitude. Table 5 summarizes equations for
wave velocity and skin depth. Typically, most fluids
and solids in the subsurface are either paramagnetic
or diamagnetic; then, velocity and skin depth equa-
tions become those listed in the second half of Table
5.
Magnetization and polarization losses add to
Ohmic conduction to render an effective conductiv-
ity. When the effective conductivity is high, the first
term of the wave equation prevails and the electro-
magnetic transient propagates in diffusional mode.

138
Table 4. Elastic Waves: Attenuation


A2  r1 
1 Attenuation    T e  ( r2 r1 )
A1  r2 
2 Ratio between the amplitude of particle motion A at radial distances r1 and r2 [m]
3 Transmission coefficient T across an interface: a function of relative impedance
4 Exponent: plane wave ς=0; cylindrical front ς=0.5; spherical front ς=1
5 Relationship between linear attenuation coefficient  [1/m], skin depth Sd [m], and damping ratio D:
1 2D Sd 1
6   Note: the skin depth in terms of wavelengths is 
Sd   2D

7 Physical processes in material attenuation

Dry - small strain: thermo-elastic relaxation
Moist/wet - small strain: viscous loss prevails
Large strain: frictional loss

Typical damping values at small strain for P- and S- waves

Material Damping D Comments
Air 2×10-4 Changes with relative humidity
Water 2×10-6 Increases with dissolved gas
Coarse soils dry 0.002~0.008 Increases when wet
saturated 0.005~0.02
Fine soils saturated 0.01~0.05
Organic soils saturated 0.01~0.05
Rocks dry 0.002~0.004
wet 0.006~0.025
Notes:  is the strain, 0' is the effective confinement, wg is the gravimetric water content. Sources: Johnston and Toksoz 1980; Yasuda
and Matsumoto (1993); Kim et al. (1991); Laird and Stokoe (1993); Santamarina and Cascante (1996); Li et al. (1998); Kokusho
(1980); Kokusho et al. (1982); Cascante and Santamarina (1996); Diaz-Rodriguez and Santamarina (2001); Kim and Novak (1981).

Table 5. Electromagnetic Waves: Velocity and Attenuation.

Electromagnetic wave propagation
2E E 2E
Maxwell’s Equations 2
    2
x t t

Electromagnetic properties
Permeability μ* Relative r=μ*/μ0 free space 0 = 4107 H/m
Permittivity * Relative κ=ε*/ε0 free space ε0= 8.85×10-12 C2N-1m-2
Conductivity σel

Velocity (non-ferromagnetic) Attenuation (non-ferromagnetic)

1 Linear attenuation α=1/Sd

V  c0
 2  c 1
1 2    Sd  0

      
 0 
2
2   1 2   
      
2   0  
In free space c0 = 1 /  0  0 =3×108 m/s  

139
Table 6. Electrical Conductivity.
Availability and mobility of hydrated ions
E E

Hydrated ion Double layer surface

Note: The effective conductivity rises as the frequency increases

Conductivity of single-phase soil components ( in S/m)

de-ionized water 10-6 organic fluids ~10-11
fresh water 10-3 most soil-forming minerals 10-15-10-7
sea water 4 (Note: some minerals are conductive)

9 Conductivity of
mixtures
el in mS/m (Annan, 1992)
water + salt = elec-  el  0.15 TDS TDS: total dissolved salts in mg/L
trolyte
Water saturation Sr= 100%
 soil  n el  1  n g S s  surface conduction.
Needs correction for tortuosity
wet soils
(Archie, 1942)
 soil  a el S rc n m
a1, m~1-2.4, c~4-5

Trend for soils ( in S/m)

 Controlled by (1- el=

soil
) S
1
0
Con-
c
trolled by
1 S
3

sa

1 1 1 
6 3
de-ion- fresh sea
0 el
ized water water
Notes: The surface conduction for kaolinite is about   10-9 Siemens. Tortuosity may reduce the electrical conductivity in clays
more than in sands. Hence, the conductivity of marine clays may be lower than the conductivity of marine sands, at the
same void ratio.
Sources: Annan (1992); Reynolds (1997); Santamarina et al. (2001). Based on an earlier compendium by Santamarina et al. (2005).

6 lists convenient semi-empirical models. Pore fluid

Electrical conductivity. The movement of hydrated
ions is responsible for electrical conduction in geo- conductivity prevails in sands and silts, while surface
conduction gains relevance in low porosity, high spe-
materials. Hydrated ions imply the presence of water,
cific surface clayey sediments saturated with low salt
dissolved salts (both cations and anions), and hy-
drated counter-ions next to charged mineral surfaces concentration water (Table 6).
Permittivity. Permittivity is a measure of polariza-
(diffuse double layer). Therefore, the electrical con-
bility. Free water molecules control the permittivity
ductivity of soils is proportional to the volumetric wa-
of soils and rocks in the MHz-to-GHz frequency
ter content, the conductivity of the pore fluid which
range; therefore, the permittivity of wet soils is pro-
increases linearly with salt concentration c (for low
portional to the volumetric water content θv=Srn. By
c<~0.2M), and the specific surface of the soil; Table

140
contrast, the orientational polarization of water is hin-
dered when water is frozen, in adsorbed layers close
to mineral surfaces and when water molecules hy-
drate ions. Table 7 summarizes semi-empirical mod-
els.
Permeability. We can assume that soils are non-fer-
romagnetic in the absence of ferromagnetic inclu-
sions. Otherwise, the volume fraction of ferromag-
netic inclusions determines the magnetic
permeability of the soil (Table 8).

Table 7. Permittivity (Relevant frequency range 1 MHz-1 GHz).

Polarization
E E

orientation of interfacial po-

polar molecules larization
Note: Permittivity increases as frequency decreases
Permittivity of single-phase soil components (radio frequencies)
water 78.5 quartz 4.2 - 5
methanol 32.6 calcite 7.7 - 8.5
most organic fluids 2 - 6 most minerals 6 - 10
Permittivity of wet soils (v=Srn)
 '  1.40  87 .6   18 .7  2 50 MHz Wensink (1993)
v v

MHz to
 '  3.03  9.3  v  146.0  v2  76.7  v3 Topp et al. (1980)
GHz
 '  3.3  41.4 v  16.0 v2 Wensink (1993)

 '  3.14  23.8 v  16.0 v2 Wang (1980)

 '  40 v  3.9  44.8  392 v  1600 v2 ~ 1GHz Selig and Mansukhani (1975)
 '  2.6  1.6 n  7.9  v  2 Based on CRI mixture model

Table 8: Magnetic Permeability.

Magnetization
H H

(a) (b) (c)

Magnetization mechanisms in ferromagnetic materials: (a) no magnetic field H=0; (b) rotation of spins within do-
mains H>0; and (c) translation of domain walls H>0.
Single materials
water, quartz, kaolinite ~0.9999 (diamagnetic)
montmorillonite, illite, granite, hematite 1.00002-1.0005 (paramagnetic)
nickel, iron > 300 (ferromagnetic)
Predictive relations
Wagner's model for spherical particles (1) '  1  3v Fe for vFe<0.2
Kaolinite with iron filings (at 10 kHz) (2) '  1  4v Fe  7 v 2Fe for vFe<0.3

Sources: (1) Göktürk et al. (1993); (2) Klein and Santamarina (2000)
Note: vFe is the volume fraction of ferromagnetic inclusions

141
5 THERMAL PROPERTIES
Table 9 schematically summarizes thermal conduction
phenomena.

Table 9: Thermal Phenomena in Soils.

Parameters
Latent heat L [kJ/kg]: Heat required for phase transformation at constant temperature.
Specific heat cp [Jkg-1K-1]: Heat required to increase the temperature of 1 kg by 1 K.
Thermal conductivity kT [Wm-1K-1]: relates the heat flux density q [J/sec/m2] to the thermal gradient
q= kT T/x (Fourier’s heat law)
Thermal diffusivity: DT=kT/ρcp.
Volumetric expansion coeff T [K-1]: Relates temperature change to strain =T

Typical values
Latent Specific Thermal con- Thermal diffu- Volumetric thermal
Material heat heat cp duct. kT sivity DT [m2sec- expansion coefficient
[kJ/kg] [Jkg-1K-1] [Wm-1K-1] 1
×10-7] T [10-6 K-1]
Quartz (single crystal) 750 12 (∥) - 6.8 (�) 45 800 (∥)-1400 (�)
Shale 630 1.56 31
Limestone 900 1.3 27 3.3
dry 800 0.15 - 0.33
Sand
water sat 2200 2-4
Water 334 4200 0.6 1.4 200 (at 293K)
Ice (0 °C) 334 2040 2.2 11.2 51 (at 273K)
Air 1000 0.024 0.21 3400 (at 293K)

Trends: General kT dry soil < kT wet soil < kT mineral

0.135 dry  64.7
Dry soils kT ,dry  (Johansen 1975)
 g  0.947  dry
Wet soils kT ,wet  kT ,dry  (1  e 0.89Sr )(kT ,sat  kT ,dry ) (Ewen and Thomas 1987)

Thermal conduction: Grain and pore-scale processes

Particle conduction kT increase with:

Contact conduction
Radiation
Particle-particle radiation
Particle-fluid conduction
Particle-fluid-part conduction
Pore fluid conduction
Pore fluid convection
 porosity n ↓
 effective stress σ' ↑
(heat transfer at contacts ↑)
 water content ↑
 quartz content ↑
 frozen water
 coarser grains

A change in temperature propagates as a diffusion
front through the soil mass, facilitated by the thermal
conductivity of the soil yet hindered by the heat con-
sumed in changing the temperature of the soil mass.
The propagating thermal front causes volume strains,
and phase changes may take place (vapourliq-
uidsolid). Consequently, the thermal properties of
soils include: thermal diffusivity DT, thermal conduc-
tivity kT, specific heat cp, latent heat L, and volumetric
thermal expansion coefficient T. Table 9 presents
definitions and typical values.
In general, we can use tabulated values and gravimet-
ric or volumetric averages to estimate thermal prop-
erties for engineering analyses, except for the thermal
conductivity (and hence diffusivity). Thermal con-
duction depends on:
 The type of mineral: quartz exhibits particularly
high thermal conductivity
 The coordination number between grains:
denser soils exhibit higher kT
 The quality of contacts: contact resistance de-
creases as stress increases (it follows a power

142
 function that resembles Hertzian contact behav- 8 LIST OF SYMBOLS
iour), and in cemented soils
 The presence of water: while the thermal con-  attenuation coefficient
ductivity of water is lower than that of most min- T volumetric thermal expansion coefficient
erals, the presence of water at contacts has a pro- ,  fitting parameters in velocity-stress relation
nounced effect on the thermal conduction across ∆ energy loss/cycle
contacts and the conductivity of the soil.  strain
* complex permittivity
0 permittivity of free space, 0 = 8.85  10-12 C2/(Nm2) =
8.85  10-12 F/m
6 CLOSING THOUGHTS: GEOPHYSICS AND  shear strain (elas elastic threshold strain)
ENGINEERING  relative permittivity
 wavelength
The previous sections identify various causal links * magnetic permeability (subscript r: relative permeability)
between geophysical properties, soil type, index 0 magnetic permeability of free space, 0 = 4  107 H/m
properties and soil behaviour. Salient causal relations  Poisson’s ratio
follow: v volumetric water content
 surface conduction
 Soil type: related to stress-dependent changes in
 mass density (fl: fluid; g: mineral that makes the grains;
the S-wave velocity, thermal conductivity, elec- sus: suspension; dry: bulk dry)
trical conductivity and permittivity. ς exponent in geometric attenuation that depends on wave
 Small strain shear stiffness (for deformation- front
based design): it is computed from S-wave ve- el electrical conductivity (el: electrolyte; fl: fluid)
locity Vs measurements using cross-hole down-  stress (': effective stress; v: vertical; h: horizontal; mean:
hole and surface wave methods. Shear stiffness mean in polarization plane)
and Vs reflect effective stress –including suc-  angular frequency
tion, extent of diagenetic cementation and/or A amplitude
freezing. Exercise caution when soils are un- a, c, m constants used in Archie’s equation
saturated as variation in saturation levels will re- B bulk stiffness (sk: soil skeleton; g: mineral that makes
sult in changes in the shear stiffness. grains; sus: suspension; fl: fluid)
Cc compression index
 Porosity: from the volumetric water content cal- Cu coefficient of uniformity
culated with electromagnetic measurements us- cbrine concentration of NaCl brine [mol/L]
ing ground penetrating radar, electrical resistiv- co speed of light in free space, co = 3108 m/s
ity or time domain reflectometry. In addition, cp specific heat [Jkg-1K-1]
electromagnetic parameters can be used to de- D damping ratio
tect time-varying water saturation, pore fluid DT thermal diffusivity
chemistry, and fluid phase (e.g., freezing). E electric field
 Other geophysical parameters mirror time vary- E Young’s modulus
F soil fraction (G: gravel; S: sand; F: fines)
ing water saturation, including P-wave velocity e void ratio (subscripts G and S denote the gravel and sand,
when Sr100%, S-wave velocity for intermedi- respectively)
ate saturations, and thermal conductivity when f frequency (r: resonant frequency)
Sr0%. G shear modulus (sk: soil skeleton)
 Thermal properties are needed for the design of Gker specific gravity of kerosene
systems such as buried cable installations and H magnetic field
thermal piles. Specific and latent heats are grav- kT thermal conductivity (sat: saturated)
L latent heat
imetric or volumetric averages of tabulated val-
LL liquid limit (ker: kerosene; DW: de-ionized water)
ues for the individual soil components. The ther- M constraint modulus
mal conductivity (and hence diffusivity) can be n porosity
measured in the lab under proper field stress and R roundness
moisture conditions. r distance from source
Sd skin depth
SE electrical sensitivity
7 ACKNOWLEDGEMENTS Sr degree of saturation
Ss specific surface
T transmission coefficient
This research is supported by the KAUST endow- TDS total dissolved salts in mg/L
ment. G. Abelskamp edited the manuscript. The au- V wave velocity (ph: phase; P: P-wave; S: S-wave; R: Ray-
thors are grateful to the organizers of the 5th Interna- leigh wave)
tional Conference on Geotechnical and Geophysical vFe volume fraction of ferromagnetic inclusions
Site Characterization. wg gravimetric water content

143
9 REFERENCES O'Konski, C. T. (1960). Electrical properties of macromolecules
V. Theory of ionic polarization in polyelectrolytes. Journal
Annan, A. P. (1992). Ground penetrating radar, Workshop of Physical Chemistry, 64, 605-618.
Notes, Sensors and Software, Mississauga. Park, J and Santamarina, J. C. (2016). Revised soil classification
Archie, G. E. (1942). The electrical resistivity log as an aid in system for coarse-fine mixtures. Journal of Geotechnical
determining some reservoir characteristics: Transactions of and Geoenvironmental Engineering, ASCE (Under review).
the American Institute of Mining, Metallurgical and Petro- Reynolds, J.M. (1997). An introduction to applied and environ-
leum Engineers, 146, 54-62. mental geophysics. John Wiley and Sons, Chichester, U.K.,
Cascante, G. and Santamarina, J. C. (1996). Interparticle contact 796 pages.
behavior and wave propagation. Journal of Geotechnical En- Richart, F. E., Hall, J. R., and Woods, R. D. (1970). Vibrations
gineering, ASCE, 122(10), 831-839. of soils and foundations, Prentice Hall, Englewood Cliffs,
Cho, G. C. and Santamarina, J. C. (2001). Unsaturated particu- 414 pages.
late materials - particle level studies. Journal of Geotech- Santamarina, J. C. and Cascante, G. (1996). Stress anisotropy
nical Geoenvironmental Engineering, ASCE, 127, 84-96. and wave propagation: a micromechanical view. Canadian
Dìaz-Rodrìguez, J.A. and Santamarina, J.C. (2001), Mexico city Geotechnical Journal, 33, 770-782.
soil behavior at different strains: observations and physical Santamarina, J. C., Klein, K. A., and Fam, M. A. (2001). Soils
interpretation. Journal of Geotechnical and Geoenvironmen- and waves. John Wiley & Sons, LTD. New York, 488 pages.
tal Engineering, ASCE, 127, 783-789. Selig, E. T. and Mansukhani, S. (1975). Relationship of soil
Ewen, J., and Thomas, H. R. (1987). The thermal probe - a new moisture to the dielectric property. Journal of the Geotech-
method and its use on an unsaturated sand. Geotechnique, nical Engineering Division, 101, 757-770.
37(1), 91-105. Topp, G. C., Davis, J. L., Bailey, W. G. and Zebchuk, W. D.
Göktürk, H. S., Fiske, T. J. and Kalyon, D. M. (1993). Electric (1980). The measurement of soil water content using a port-
and magnetic properties of a thermoplastic elastomer incor- able TDR hand probe. Canadian Journal of Soil Science, 64,
porated with ferromagnetic powders. IEEE Transactions on 313-321.
Magnetics, 29, 4170-4176. Wang, J. R. (1980). The dielectric properties of soil-water mix-
Hardin, B. O. and Drnevich, V. P. (1972). Shear modulus and tures at microwave frequencies. Radio Science, 15, 977-985.
damping in soils: measurement and parameter effects. Jour- Wensink, W. A. (1993). Dielectric properties of wet soils in the
nal of Soil Mechanics Foundation Division, ASCE, 98, 603- frequency range 1-3000 MHz. Geophysical Prospecting, 41,
624. 671-696.
Jang, J. and Santamarina, J. C. (2016a). Fines classification Yasuda, N. and Matsumoto, N. (1993). Dynamic deformation
based on sensitivity to pore-fluid chemistry. Journal of Ge- characteristics of sand and rockfill materials. Canadian Ge-
otechnical and Geoenvironmental Engineering, ASCE, 142 otechnical Journal, 30, 747-757.
(4), p.06015018.
Jang, J. and Santamarina, J. C. (2016b). Closure of “Fines clas-
sification based on sensitivity to pore-fluid chemistry”. Jour-
nal of Geotechnical and Geoenvironmental Engineering,
ASCE (in progress).
Johansen, T. A. (1975). Thermal conductivity of soils. US Army
Corps of Engineers, Trondheim, Hanover, New Hampshire.
Johnston, D.H. and Toksöz, M.N. (1980). Ultrasonic P and S
wave attenuation in dry and saturated rocks under pressure.
Journal of Geophysical Research: Solid Earth, 85(B2), 925-
936.
Kim, T.C. and Novak, M. (1981). Dynamic properties of some
cohesive soils of Ontario. Canadian Geotechnical Journal,
18, 371-389.
Kim, D. S., Stokoe, K. H. and Hudson, W. R. (1991). Deforma-
tional characteristics of soils at small to intermediate strains
from cyclic tests. Report 1177-3, Center for Transportation
Research, Bureau of Engineering Research, the University of
Texas Austin, 142 pages.
Klein, K. and J. C. Santamarina. (2000). Ferromagnetic inclu-
sions in geomaterials – implications. Geotechnical Journal,
ASCE, 162, 167-179.
Kokusho, T. (1980). Cyclic triaxial test of dynamic soil proper-
ties for wide strain range. Soils and Foundations, 20, 45-60.
Kokusho, T., Yoshida, Y. and Esashi, Y. (1982). Dynamic prop-
erties of soft clay for wide strain range. Soils and Founda-
tions, 22, 1-18.
Laird, J. P. and Stokoe, K. H. (1993). Dynamic properties of
remolded and undisturbed soil samples tested at high confin-
ing pressures. Geotechnical Engineering Report GR93-6,
Electrical Power Research Institute, Palo Alto, Calif.
Li, X. S., Yang, W. L., Shen, C. K., and Wang, W. C. (1998).
Energy-injecting virtual mass resonant column system. Jour-
nal of Geotechnical and Geoenvironmental Engineering,
ASCE, 124, 428-438.

144